Research On Composition Structure Selection Of Synthetic Image In Story Visualization

With the development of technology，various of techniques are shown in front ofpeople more and more intuitively, changing the way we live and communicate quietly.Take storytelling as an example, we change the way of telling story from language,words to pictures. We can improve the efficiency of communication by using picturesgotten from story visualization as an auxiliary way. By referencing others’ work inthis field, we can select sub graph having more better composition structure usingimage segmentation and image synthesis. The main contribution of this work can besummarized follows:Firstly, we use an image retrieval method based on semantic information tosearch pictures in the system. After searching, we use image segmentation and imagesynthesis to get the picture that fits our needs better.Then, we can select sub graph having much better composition structure. In thisprocess, although the new pictures can express the correct meaning of user, they arenot able to give people the feeling of visual beauty. We use the one-third principle anddiagonal line principle, which meet the principle of people’s aesthetic view, to selectsub graph of the synthetic image.Finally, we try to add grammatical constraint to ensure that the synthetic imagemeets the user’s need through the process of image synthesis. We know the relativeposition relationship of the objects in the image according to grammatical constraint,which ensure that the synthesis result matches the input sentence. Therefore, the finalresult will be logical, rather than a meaningless images.